Intel Empowers Builders With OneAPI 2022 Toolkits

Intel launched oneAPI 2022 toolkits. Newly enhanced toolkits make bigger cross-architecture options to supply builders…

Intel launched oneAPI 2022 toolkits. Newly enhanced toolkits make bigger cross-architecture options to supply builders larger software and architectural option to boost up computing.

About New Features: New features come with the sector’s first unified compiler imposing C++, SYCL and Fortran, information parallel Python for CPUs and GPUs, complicated accelerator functionality modeling and tuning, and function acceleration for AI and ray tracing visualization workloads. The oneAPI cross-architecture programming fashion supplies builders with equipment that purpose to  give a boost to the productiveness and speed of code building when development cross-architecture packages.

Why It Issues: Consistent with an Evans Information survey, 40% of builders goal heterogeneous programs that use multiple form of processor, processor core or coprocessor.1 The tempo of innovation is handiest accelerating with cross-architecture computing pushed ahead with oneAPI throughout heterogeneous materials of CPUs, GPUs, FPGAs and different accelerators — all of which require tool to bind those components into a whole answer. With oneAPI, builders have the liberty to make a choice the most productive {hardware} for a given answer with out the commercial and technical burdens of proprietary programming fashions.

Extra Main points: The 2022 Intel® oneAPI toolkits ship functionality and productiveness thru a whole set of complicated equipment together with compilers, libraries, pre-optimized frameworks, analyzers and debuggers. There are greater than 900 new and enhanced options added during the last 12 months that reinforce each and every instrument within the foundational and domain-specific toolkits. They’re now to be had to obtain or use within the Intel® DevCloud free of charge. Highlights come with:

Move-architecture programming

  • Intel created the sector’s first unified compiler imposing C++, SYCL and Fortran for CPUs and GPUs using a commonplace LLVM backend.
  • Speeded up compute on CPUs and GPUs for Python, these days’s hottest programming language.
  • The Intel® DPC++ Compatibility Instrument used to be progressed to routinely migrate 90% to 95% of CUDA code to SYCL/DPC++. 2

Efficiency on the most recent {hardware}

  • {Hardware} fortify – Intel oneAPI Toolkits are optimized to allow complicated options of the most recent and upcoming new {hardware}, together with twelfth Gen Intel® Core™ processors with AVX-VNNI, Subsequent Gen Intel® Xeon® Scalable processors, codenamed Sapphire Rapids with Intel® Complex Matrix Extension (Intel® AMX), and upcoming Xe shopper and knowledge middle GPUs.

AI functionality optimizations

  • Deep studying framework functionality is speeded up as much as 10 occasions over previous variations with the most recent Intel® Optimization for TensorFlow and Intel® Optimization for PyTorch.3
  • New Intel® Extension for Scikit-learn accelerates device studying algorithms greater than 100 occasions on Intel CPUs over the inventory open supply model.4
  • Offered Intel® Neural Compressor to reach greater inference functionality thru post-training optimization tactics throughout a couple of deep studying frameworks.

Complex equipment for building productiveness

  • Professional functionality research for CPUs and accelerators – Intel® VTune™ Profiler’s flame graph show is helping give a boost to the power to visualise functionality sizzling spots. Intel® Consultant’s accelerator functionality modeling permits builders to estimate functionality advantages of offloading to a GPU prior to making code adjustments.
  • Complex ray tracing – New options together with cone telemetry, auxiliary function denoising and FP16 fortify supply extra tough shapes and shorten rendering occasions, making improvements to total rendering functionality. Fortify for upcoming Intel Xe GPUs in addition to real-time denoising additional give a boost to industry-leading, ultimate body, production-quality rendering.
  • Expanded building setting fortify together with deeper Microsoft Visible Studio Code integration, fortify for Microsoft Visible Studio 2022, and Microsoft WSL2 for Linux building on Home windows.
See also  Intel and MediaTek Shape Foundry Partnership

About Intel oneAPI Toolkits: Intel oneAPI toolkits ship the equipment had to successfully broaden high-performance packages and answers throughout a lot of Intel architectures (CPUs, GPUs, FPGAs). Its set of complementary toolkits — a base toolkit and domain-specific add-ons — simplify programming and lend a hand builders give a boost to productiveness, ship uncompromised functionality and boost up innovation. The toolkits come with complicated optimized compilers, libraries, frameworks and research equipment for performant heterogeneous computing. They put into effect the oneAPI specs and paintings along complementary optimization equipment.

About oneAPI: oneAPI is an open, unified and cross-architecture programming fashion for CPUs, GPUs, FPGAs and different accelerators. In response to requirements, the programming fashion simplifies tool building and delivers uncompromised functionality for speeded up computing with out proprietary lock-in, whilst enabling the combination of legacy code.

Supply: Intel